Plant uptake of sulphur as related to changes in the HI-reducible and total sulphur fractions in soil

Although considerable progress has been made in relating extractable soil S to plant S availability, most of these studies determined the extractable soil S at the beginning of the experiment to use as an index of soil S status. This bears little or no relationship to the S taken up by plants during the entire growing season. The present study investigates the changes in extractable soil S with time and relates these to changes in plant S uptake. Six soils with different long-term fertiliser histories (0, 21, 42 kg of S as superphosphate ha–1 applied since 1952) and animal camping treatments (camp and non-camp) were used in two pot systems (with and without plants). Carrier-free 35SO4–S was added to the soils, to provide the information on the transformations of recently added S between the different extractable S forms in soils and whether these transformations could predict plant-available S. The soils were pre-conditioned and then transferred to the glasshouse, where one set of pots were planted with perennial ryegrass (Lolium perenne L.) while the other set was left uncropped. Periodic plant harvests and soil samplings at four weekly intervals were conducted over a period of 20 weeks to determine plant S uptake and amounts of extractable soil S and 35S forms using five extractants. Same extractions of soil S and 35S were conducted for the initial soils. Results showed that HI-reducible and total soil S extracted by CaCl2, KH2PO4 and by KCl at 40°C were utilised significantly by plants but not those extracted by NaHCO3 and NaOH extractants. However, after the 8th week, plants continued to take up S even though levels of S extracted from the soil by CaCl2, KH2PO4 and by KCl at 40°C remained low and unchanged. These results suggest that soil S taken up by plants after the 8th week period originated directly from the mineralisation of soil organic S from S pools other than those present in the extractable soil S forms. Similar results were shown by 35S data, thereby confirming the complexity of determining plant S availability based on soil S extraction methods.